This is a fascinating and revealing discussion. It was interesting to see how little reverb was used in the Ed Sheeran. Reverb can easily be over-used as it is flattering to the singer (singers always ask for more reverb!) but it also makes the vocals seem more distant. It is clear that delay has been used as well as compression. There is very little variety of vocal volume so when the voice goes low it is practically as loud as when he is singing at higher pitches and obviously projecting more. Therefore something has been applied to level the vocals. You also have to be careful with levelling as it can take out the expression and variety. There might also be some auto-double tracking applied in the Ed Sheeran example.

Thanks for the thread as it made me think about my own vocal processing methods and maybe revise them a little.

www.tonycliffmusic.com
http://tonycliff.bandcamp.com